INDEX

Explanations

words related to uncovering or discovering something

oai_token-act-pair · gpt-3.5-turbo

variations of the word "covered" and its related forms in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

wcsstore

-0.75

¦

-0.66

primary

-0.65

sq

-0.63

DEV

-0.62

Built

-0.62

MF

-0.60

¾

-0.60

Ĩ

-0.60

Ak

-0.59

POSITIVE LOGITS

vironment

0.91

overed

0.89

uyomi

0.88

xtap

0.86

ertodd

0.84

rition

0.77

olia

0.74

 Canaver

0.74

 Collider

0.74

hower

0.74

Activations Density 0.034%

words related to uncovering or discovering something

variations of the word "covered" and its related forms in various contexts

No Comments

No Known Activations